BenchTemp: A General Benchmark for Evaluating Temporal Graph Neural Networks
This work addresses the need for consistent and diverse benchmarking in the temporal graph neural network community, though it is incremental as it builds on existing TGNN methods rather than introducing new models.
The authors tackled the problem of inconsistent and limited evaluations for temporal graph neural networks (TGNNs) by proposing BenchTemp, a general benchmark that provides standardized datasets and pipelines, enabling comprehensive comparisons of TGNN models on tasks like link prediction and node classification, with results showing improved fairness and efficiency in evaluations.
To handle graphs in which features or connectivities are evolving over time, a series of temporal graph neural networks (TGNNs) have been proposed. Despite the success of these TGNNs, the previous TGNN evaluations reveal several limitations regarding four critical issues: 1) inconsistent datasets, 2) inconsistent evaluation pipelines, 3) lacking workload diversity, and 4) lacking efficient comparison. Overall, there lacks an empirical study that puts TGNN models onto the same ground and compares them comprehensively. To this end, we propose BenchTemp, a general benchmark for evaluating TGNN models on various workloads. BenchTemp provides a set of benchmark datasets so that different TGNN models can be fairly compared. Further, BenchTemp engineers a standard pipeline that unifies the TGNN evaluation. With BenchTemp, we extensively compare the representative TGNN models on different tasks (e.g., link prediction and node classification) and settings (transductive and inductive), w.r.t. both effectiveness and efficiency metrics. We have made BenchTemp publicly available at https://github.com/qianghuangwhu/benchtemp.